Chetan Edupalli

Implemented autonomous navigation, multi-sensor fusion mapping, and perception pipelines, deploying them on Edge hardware to decrease processing latency by 20%. Co-led a multidisciplinary team of 10+ engineering students to secure 1st place in the Autonomous task at both the Australian Rover Challenge and the University Rover Challenge.

Designed and built an Agentic AI dual-engine backend using FastAPI with strict JSON schema to eliminate UI hallucinations. Developed context-aware prompt engineering to synthesize behavioral telemetry data with LLM reasoning to predict stress thresholds with an 88% recall rate.

Researched Multi-Agent Path Finding (MAPF) optimization for sub-optimal LNS-based solvers under the supervision of Dr. Daniel Harabor. Benchmarked algorithm execution speeds and scaled performance tests across 1,000+ agents, achieving competitive runtime results compared to existing algorithms like LNS2 and DropLNS.

Scaled the organization's financial backing from 0 to 3 major sponsors within 6 months, securing vital funding for student hackathons.

Developed, trained, and optimized deep learning transformer models to translate American Sign Language to text in real-time. Achieved a 94.5% translation accuracy rate while reducing model inference time to under 45ms for seamless live deployment.

Engineered and optimized large-scale, cloud-based data pipelines to handle terabytes of data, utilising AWS analytics tools to streamline query performance, improving data processing efficiency by 15%.

Pursuing a Master of Artificial Intelligence, with expected completion in July 2026.

Completed a Bachelor of Technology in Computer Science and Engineering from September 2020 to June 2024.